Hibernate.orgCommunity Documentation

Chapter 6. Manual index changes

6.1. Adding instances to the index
6.2. Deleting instances from the index
6.3. Rebuilding the whole index
6.3.1. Using flushToIndexes()
6.3.2. Using a MassIndexer
6.3.3. Useful parameters for batch indexing

As Hibernate core applies changes to the Database, Hibernate Search detects these changes and will update the index automatically (unless the EventListeners are disabled). Sometimes changes are made to the database without using Hibernate, as when backup is restored or your data is otherwise affected; for these cases Hibernate Search exposes the Manual Index APIs to explicitly update or remove a single entity from the index, or rebuild the index for the whole database, or remove all references to a specific type.

All these methods affect the Lucene Index only, no changes are applied to the Database.

6.1. Adding instances to the index

Using FullTextSession.index(T entity) you can directly add or update a specific object instance to the index. If this entity was already indexed, then the index will be updated. Changes to the index are only applied at transaction commit.

Example 6.1. Indexing an entity via FullTextSession.index(T entity)

FullTextSession fullTextSession = Search.getFullTextSession(session);

Transaction tx = fullTextSession.beginTransaction();

Object customer = fullTextSession.load( Customer.class, 8 );
fullTextSession.index(customer);

tx.commit(); //index only updated at commit time

In case you want to add all instances for a type, or for all indexed types, the recommended approach is to use a MassIndexer: see Section 6.3.2, “Using a MassIndexer” for more details.

The method FullTextSession.index(T entity) is considered an explicit indexing operation, so any registered EntityIndexingInterceptor won't be applied in this case. For more information on EntityIndexingInterceptor see Section 4.5, “Conditional indexing: to index or not based on entity state”.

6.2. Deleting instances from the index

It is equally possible to remove an entity or all entities of a given type from a Lucene index without the need to physically remove them from the database. This operation is named purging and is also done through the FullTextSession.

Example 6.2. Purging a specific instance of an entity from the index

FullTextSession fullTextSession = Search.getFullTextSession(session);

Transaction tx = fullTextSession.beginTransaction();

for (Customer customer : customers) {

    fullTextSession.purge( Customer.class, customer.getId() );

}

tx.commit(); //index is updated at commit time

Purging will remove the entity with the given id from the Lucene index but will not touch the database.

If you need to remove all entities of a given type, you can use the purgeAll method. This operation removes all entities of the type passed as a parameter as well as all its subtypes.

Example 6.3. Purging all instances of an entity from the index

FullTextSession fullTextSession = Search.getFullTextSession(session);

Transaction tx = fullTextSession.beginTransaction();
fullTextSession.purgeAll( Customer.class );

//optionally optimize the index

//fullTextSession.getSearchFactory().optimize( Customer.class );

tx.commit(); //index changes are applied at commit time    

As in the previous example, it is suggested to optimize the index after many purge operation to actually free the used space.

As is the case with method FullTextSession.index(T entity), also purge and purgeAll are considered explicit indexinging operations: any registered EntityIndexingInterceptor won't be applied. For more information on EntityIndexingInterceptor see Section 4.5, “Conditional indexing: to index or not based on entity state”.

Note

Methods index, purge and purgeAll are available on FullTextEntityManager as well.

Note

All manual indexing methods (index, purge and purgeAll) only affect the index, not the database, nevertheless they are transactional and as such they won't be applied until the transaction is successfully committed, or you make use of flushToIndexes.

6.3. Rebuilding the whole index

If you change the entity mapping to the index, chances are that the whole Index needs to be updated; For example if you decide to index a an existing field using a different analyzer you'll need to rebuild the index for affected types. Also if the Database is replaced (like restored from a backup, imported from a legacy system) you'll want to be able to rebuild the index from existing data. Hibernate Search provides two main strategies to choose from:

This strategy consists in removing the existing index and then adding all entities back to the index using FullTextSession.purgeAll() and FullTextSession.index(), however there are some memory and efficiency contraints. For maximum efficiency Hibernate Search batches index operations and executes them at commit time. If you expect to index a lot of data you need to be careful about memory consumption since all documents are kept in a queue until the transaction commit. You can potentially face an OutOfMemoryException if you don't empty the queue periodically: to do this you can use fullTextSession.flushToIndexes(). Every time fullTextSession.flushToIndexes() is called (or if the transaction is committed), the batch queue is processed applying all index changes. Be aware that, once flushed, the changes cannot be rolled back.

Example 6.4. Index rebuilding using index() and flushToIndexes()

fullTextSession.setFlushMode(FlushMode.MANUAL);

fullTextSession.setCacheMode(CacheMode.IGNORE);

transaction = fullTextSession.beginTransaction();

//Scrollable results will avoid loading too many objects in memory

ScrollableResults results = fullTextSession.createCriteria( Email.class )

    .setFetchSize(BATCH_SIZE)

    .scroll( ScrollMode.FORWARD_ONLY );

int index = 0;

while( results.next() ) {

    index++;

    fullTextSession.index( results.get(0) ); //index each element

    if (index % BATCH_SIZE == 0) {

        fullTextSession.flushToIndexes(); //apply changes to indexes

        fullTextSession.clear(); //free memory since the queue is processed

    }

}

transaction.commit();

Try to use a batch size that guarantees that your application will not run out of memory: with a bigger batch size objects are fetched faster from database but more memory is needed.

Hibernate Search's MassIndexer uses several parallel threads to rebuild the index; you can optionally select which entities need to be reloaded or have it reindex all entities. This approach is optimized for best performance but requires to set the application in maintenance mode: making queries to the index is not recommended when a MassIndexer is busy.

Example 6.5. Index rebuilding using a MassIndexer

fullTextSession.createIndexer().startAndWait();

This will rebuild the index, deleting it and then reloading all entities from the database. Although it's simple to use, some tweaking is recommended to speed up the process: there are several parameters configurable.

Warning

During the progress of a MassIndexer the content of the index is undefined! If a query is performed while the MassIndexer is working most likely some results will be missing.

Example 6.6. Using a tuned MassIndexer

fullTextSession

 .createIndexer( User.class )

 .batchSizeToLoadObjects( 25 )

 .cacheMode( CacheMode.NORMAL )

 .threadsToLoadObjects( 12 )

 .idFetchSize( 150 )

 .progressMonitor( monitor ) //a MassIndexerProgressMonitor implementation

 .startAndWait();

This will rebuild the index of all User instances (and subtypes), and will create 12 parallel threads to load the User instances using batches of 25 objects per query; these same 12 threads will also need to process indexed embedded relations and custom FieldBridges or ClassBridges, to finally output a Lucene document. In this conversion process these threads are likely going to need to trigger lazyloading of additional attributes, so you will probably need a high number of threads working in parallel. The number of threads working on actual index writing is defined by the backend configuration of each index. See the option worker.thread_pool.size in Table 3.3, “Execution configuration”.

As of Hibernate Search 4.4.0, instead of indexing all the types in parallel, the MassIndexer is configured by default to index only one type in parallel. It prevents resource exhaustion especially database connections and usually does not slow down the indexing. You can however configure this behavior using MassIndexer.typesToIndexInParallel(int threadsToIndexObjects):

Example 6.7. Configuring the MassIndexer to index several types in parallel

fullTextSession

 .createIndexer( User.class, Customer.class )

 .typesToIndexInParallel( 2 )

 .batchSizeToLoadObjects( 25 )

 .cacheMode( CacheMode.NORMAL )

 .threadsToLoadObjects( 5 )

 .idFetchSize( 150 )

 .progressMonitor( monitor ) //a MassIndexerProgressMonitor implementation

 .startAndWait();

Generally we suggest to leave cacheMode to CacheMode.IGNORE (the default), as in most reindexing situations the cache will be a useless additional overhead; it might be useful to enable some other CacheMode depending on your data: it could increase performance if the main entity is relating to enum-like data included in the index.

Note

The MassIndexer was designed for speed and is unaware of transactions, so there is no need to begin one or committing. Also because it is not transactional it is not recommended to let users use the system during its processing, as it is unlikely people will be able to find results and the system load might be too high anyway.

The MassIndexer was designed to finish the reindexing task as quickly as possible, but this requires a bit of care in its configuration to behave fairly with your server resources.

There is a simple formula to understand how the different options applied to the MassIndexer affect the number of used worker threads and connections: each thread will require a JDBC connection.

threads = typesToIndexInParallel * (threadsToLoadObjects + 1);
required JDBC connections = threads;

Let's see some suggestions for a roughly sane tuning starting point:

  1. Option typesToIndexInParallel should probably be a low value, like 1 or 2, depending on how much of your CPUs have spare cycles and how slow a database round trip will be.

  2. Before tuning a parallel run, experiment with options to tune your primary indexed entities in isolation.

  3. Making threadsToLoadObjects higher increases the preloading rate for the picked entities from the database, but also increases memory usage and the pressure on the threads working on subsequent indexing.

  4. Increasing parallelism usually helps as the bottleneck usually is the latency to the database connection: it's probably worth it to experiment with values significantly higher than the number of actual cores available, but make sure your database can handle all the multiple requests.

  5. This advice might not apply to you: always measure the effects! We're providing this as a means to help you understand how these options are related.

Warning

Running the MassIndexer with many threads will require many connections to the database. If you don't have a sufficiently large connection pool, the MassIndexer itself and/or your other applications could starve being unable to serve other requests: make sure you size your connection pool accordingly to the options as explained in the above paragraph.

Tip

The "sweet spot" of number of threads to achieve best performance is highly dependent on your overall architecture, database design and even data values. All internal thread groups have meaningful names so they should be easily identified with most diagnostic tools, including simply threaddumps.

The provided MassIndexer is quite general purpose, and while we believe it's a robust approach, you might be able to squeeze some better performance by writing a custom implementation. To run your own MassIndexer instead of using the one shipped with Hibernate Search you have to:

  1. create an implementation of the org.hibernate.search.spi.MassIndexerFactory interface;

  2. set the property hibernate.search.massindexer.factoryclass with the qualified class name of the factory implementation.

Example 6.8. Custom MassIndexerFactory example

package org.myproject

import org.hibernate.search.spi.MassIndexerFactory

[...]


public class CustomIndexerFactory implements MassIndexerFactory {


  public void initialize(Properties properties) {

  }


  public MassIndexer createMassIndexer(...) {

    return new CustomIndexer();

  }


}       
hibernate.search.massindexer.factoryclass = org.myproject.CustomIndexerFactory

Other parameters which affect indexing time and memory consumption are:

Previous versions also had a max_field_length but this was removed from Lucene, it's possible to obtain a similar effect by using a LimitTokenCountAnalyzer.

All .indexwriter parameters are Lucene specific and Hibernate Search is just passing these parameters through - see Section 3.7.1, “Tuning indexing performance” for more details.

The MassIndexer uses a forward only scrollable result to iterate on the primary keys to be loaded, but MySQL's JDBC driver will load all values in memory; to avoid this "optimisation" set idFetchSize to Integer.MIN_VALUE.